A defense attorney who is appealing his client’s murder conviction has been reprimanded by New Mexico’s highest court after submitting a legal brief that included fabricated police testimony and invented witnesses, generated using the AI tool ChatGPT. The New Mexico Supreme Court ruled that the attorney, Stephen Aarons, should be fined $5,000 and held in contempt of court for failing to verify the accuracy of the document he filed. The court described the brief as containing “false testimony from wholly fabricated witnesses” and criticized Aarons for showing a “lack of remorse and a lack of concern for his client.” The case has now been referred to an attorney disciplinary board for further investigation. Aarons claimed in a statement that he used ChatGPT to summarize the trial proceedings when he took on the appeal last year. He admitted he did not fully understand how AI tools can sometimes generate false or misleading information, a phenomenon known as “hallucination.” Aarons expressed remorse for the mistake and said he hoped the disciplinary board would consider the error as an honest one. He also warned that the incident serves as a lesson for professionals who use AI, which he described as powerful but “sometimes unstable.” This case is part of a broader trend in which judges across the U.S. have disciplined lawyers for submitting court documents created or significantly influenced by AI tools without proper verification. Some of these cases involve AI-generated errors, such as made-up legal citations or misquotations of the law. Aarons’ case appears to be one of the most serious, as it involved fabricated witness testimony in a criminal appeal. The appeal is related to Oscar Renee Sandoval, who was convicted of murdering his children’s mother and is serving a life sentence. The appeal is still ongoing and has been assigned to a public defender, Kim Chavez Cook, who declined to comment on the matter. The New Mexico Supreme Court recently asked Aarons to explain how the fabricated details, including claims about the shooter wearing dark pants and a white shirt, ended up in his brief. Aarons explained that he fed a computer-generated transcript and other case materials to ChatGPT, expecting it to produce a “bulletproof summary.” The court’s justices expressed skepticism, with Justice Shannon Bacon questioning whether Aarons was aware of the daily media coverage on AI-related issues. “The problem with lawyers relying on AI hallucinations is an 'above-the-fold' story every single day,” she said, highlighting the growing concern over the misuse of AI in legal contexts.